TransFuser: Imitation With Transformer-Based Sensor Fusion for Autonomous Driving.

How should we integrate representations from complementary sensors for autonomous driving? Geometry-based fusion has shown promise for perception (e.g., object detection, motion forecasting). However, in the context of end-to-end driving, we find that imitation learning based on existing sensor fusion methods underperforms in complex driving scenarios with a high density of dynamic agents. Therefore, we propose TransFuser, a mechanism to integrate image and LiDAR representations using self-attention. Our approach uses transformer modules at multiple resolutions to fuse perspective view and bird's eye view feature maps. We experimentally validate its efficacy on a challenging new benchmark with long routes and dense traffic, as well as the official leaderboard of the CARLA urban driving simulator. At the time of submission, TransFuser outperforms all prior work on the CARLA leaderboard in terms of driving score by a large margin. Compared to geometry-based fusion, TransFuser reduces the average collisions per kilometer by 48%.

Impurity reveals distinct operational phases in quantum thermodynamic cycles.

We analyze the effect of impurity on the work output and efficiency of quantum Otto and quantum Carnot heat cycles, modeled as a single quantum particle in an infinite square well potential, which is the working substance. We solve this quantum mechanical system perturbatively up to first and second order in strength of the impurity for strong- and weak-coupling regimes, respectively. We derive the analytical expressions of work and efficiency for the strong-coupling regime to the first order in the strength parameter. The threshold value of the strength parameter in weak coupling is obtained up to which the numerical result agrees with the perturbative result for a repulsive and attractive impurity. To our surprise, an embedded impurity unlocks new operational phases in the system, such as a quantum heat engine, quantum refrigerator, and quantum cold pump. In addition, the efficiency of the quantum Otto heat engine is seen to reach Carnot efficiency for some parameter regimes. The cooling power and coefficient of performance of the quantum refrigerator and quantum cold pump are nontrivially affected by the impurity.

Utility of non-invasive haemoglobin monitoring in oncosurgery patients.

BACKGROUND AND AIMS: Oncosurgeries may incur massive blood loss demanding frequent blood sampling to assess blood loss and the need for intraoperative blood transfusions. Accuracy of non-invasive spectrophotometric haemoglobin (hereafter to be referred as SpHb) monitoring has been studied in various perioperative settings. The intraoperative use of Radical-7®, Masimo Corp., (Radical-7®) for SpHb monitoring may be useful during cancer surgery. The aim of this study is to evaluate the intraoperative utility of SpHb monitoring by the Radical-7® to guide intraoperative transfusion in oncosurgeries. METHODS: Fifty adult patients, undergoing oncosurgery with anticipated blood loss of more than 20% of blood volume, were selected. Continuous SpHb monitoring was performed intraoperatively and blood transfusion was based on SpHb values. Simultaneous laboratory haemoglobin (LabHb) samples were taken for validation. The accuracy of intraoperative blood transfusions based on SpHb was analysed using Error Grid Analysis. Paired measurements of SpHb and LabHb were compared using Bland-Altman plot analysis. RESULTS: There were 66 paired data points for blood transfusion from fifty patients with a correlation of 73% (P < 0.001) between SpHb and LabHb. In the Bland-Altman analysis, the bias was - 0.313 g/dl with ~ 95% of values within the limits of agreement of 1.81 g/dl to -2.44 g/dl. In the Error Grid Analysis, most data points were in the least error zone (Zone A). CONCLUSION: The Radical-7® has the advantage of providing SpHb value continuously to take prompt decision regarding blood transfusion intraoperatively.

Conversion of rice straw to monomeric phenols under supercritical methanol and ethanol.

Hydrothermal liquefaction of rice straw has been carried out using various organic solvents (CH3OH, C2H5OH) at different temperatures (250, 280 and 300 °C) and residence times (15, 30 and 60 min) to understand the effect of solvent and various reaction parameters on product distribution. Maximum liquid product yield (47.52 wt%) was observed using ethanol at 300 °C and 15 min reaction time. FTIR and NMR ((1)H and (13)C) of liquid product indicate that lignin in rice straw was converted to various monomeric phenols. GC-MS of the liquid product showed the presence of various phenol and guaiacol derivatives. Main compounds observed in liquid product were phenol, 4-ethylphenol, 4-ethyl-2-methoxyphenol (4-ethylguaiacol), 2,6-dimethoxyphenol (syringol), 2-isopropyl-5-methylphenol (thymol). Powder XRD and SEM of bio-residue showed that rice straw was decomposed to low molecular weight monomeric phenols.

Catalytic hydrothermal liquefaction of water hyacinth.

Thermal and catalytic hydrothermal liquefaction of water hyacinth was performed at temperatures from 250 to 300 °C under various water hyacinth:H2O ratio of 1:3, 1:6 and 1:12. Reactions were also carried out under various residence times (15-60 min) as well as catalytic conditions (KOH and K2CO3). The use of alkaline catalysts significantly increased the bio-oil yield. Maximum bio-oil yield (23 wt%) comprising of bio-oil1 and bio-oil2 as well as conversion (89%) were observed with 1N KOH solution. (1)H NMR and (13)C NMR data showed that both bio-oil1 and bio-oil2 have high aliphatic carbon content. FTIR of bio-residue indicated that the usage of alkaline catalyst resulted in bio-residue samples with lesser oxygen functionality indicating that catalyst has a marked effect on nature of the bio-residue and helps to decompose biomass to a greater extent compared to thermal case.

Hydrothermal conversion of lignin to substituted phenols and aromatic ethers.

Hydrothermal liquefaction of lignin was performed using methanol and ethanol at various temperatures (200, 250 and 280°C) and residence times of 15, 30 and 45min. Maximum liquid product yield (85%) was observed at 200°C and 15min residence time using methanol. Increase in temperature was seen to decrease the liquid products yield. With increase in residence time, liquid yields first increased and then decreased. FTIR and (1)H NMR showed the presence of substituted phenols and aromatic ethers in liquid products and breakage of ß-O-4 or/and α-O-4 ether bonds present in lignin during hydrothermal liquefaction was confirmed through FTIR of bio-residue. In comparison to the existing literature information, higher lignin conversion to liquid products and maximum carbon conversion (72%) was achieved in this study.